Correlates of religious, supernatural and psychosocial causal beliefs about depression among Latino immigrants in primary care

Abstract

The purpose of this cross-sectional study was to explore causal attributions about depression and to identify psychosocial factors associated with these beliefs among Latino immigrants. We interviewed 177 primary care patients with instruments to assess causal beliefs, depressive and somatic symptoms, ethnic identity and stigma. An exploratory factor analysis of the Causal Beliefs scale yielded three factors, “Balance,” “Psychosocial” and “Malevolent Spirituality/Transgressions” that were used as dependent variables in multivariate analyses. Depressive symptoms, age, country of origin and religiosity were significantly associated with particular factors of causal beliefs. Those with higher education were most likely to endorse psychosocial causal beliefs. Stigma pertained to causal beliefs related to “malevolent forces” and “personal transgressions.” In conclusion, psychosocial and religious explanations of illness were strongly endorsed by these Latino immigrants, indicating a dual system of Western-medicine and traditional beliefs. These results suggest culturally-specific interventions for improving health knowledge and communication with patients about depression.

Keywords:

• Hispanics/Latinos
• immigrants
• depression
• primary care
• causal attributions
• health beliefs
• stigma

Acknowledgements

This research was partially supported by a grant from Sigma Theta Tau International Honor Society of Nursing, Indianapolis, IN, USA. Special thanks to Joshua Farkovits, owner, Privileged Care, Inc, Tatiana Castrillon, Monica Diaz, Enrique Malamud, MD, Marta Miguez, Jasmine Remos, Gabriela Garcia Rosas, MD, Romana Suero, MA, and Johanna Tudino. Thanks to Daniel Zelterman for his consultations in categorical analysis.

Notes

Note

1. Using a Bonferroni correction for the total number of variables in the bivariate equations would yield a significant association at the.0027 level, however, this stringency is generally applied to a hypothesis driven model with one outcome of interest, rather than an exploratory model. Therefore, a less conservative p < 0.1 is appropriate in this case (Zelterman, personal communication, 15 January 2008).